Americans. They were the sons of managers, proprietors, and white-collar workers, and they belonged to families in middle to high income groups who lived in good neighborhoods. In contrast, the group who had not completed college were hired as skilled craftsmen and later advanced to managerial status. The sons and grandsons of immigrants, this group had fathers who were skilled or unskilled laborers with an average of grammar school education or less. They had grown up in families of low income in modest to substandard neighborhoods. The second group (presumably less well prepared for the demands and expectations of managerial status) shared a significantly greater number of illnesses of all sorts than did the first group.
The findings of a study by Cassel and Tyroler (47) lead to similar conclusions. They studied two groups of rural mountaineers who worked in a factory. One group was composed of individuals who were the first of their family ever to engage in industrial work; the second differed from the first in that they were children of previous workers in the factory-but they worked in the same factory, came from the same mountain coves, were of the same age, and did the same work for the same wages as the first group. The study was undertaken to test the hypothesis that the second group, by virtue of their prior familial experience, would be better prepared for the expectations and demands of industrial living then would the first group and would therefore exhibit fewer signs of ill health. Health status was measured by responses to the Cornell Medical Index and by various indices of sick absenteeism. As predicted, the sons of previous factory workers had lower Cornell Medical Index scores (fewer symptoms) and lower rates of sick absenteeism after the initial few years of service, at each age, than had the "first generation" workers.
Perhaps even more convincing was the study conducted by Haenszel and his associates (48) on death rates from lung cancer in the United States. These investigators discovered that death rates from lung cancer, when controlled for degree of cigarette smoking, were considerably higher in the farm-born who had migrated to cities than they were in lifelong urban dwellers. The study had been designed initially to attempt to quantify the importance of length of exposure to atmospheric pollution of the cities, but despite the lifetime of exposure, urban dwellers had apparently "adapted" better to the effects of such atmospheric pollution than had the migrants.
Group Membership and Health
A further important concept in developing a useful formulation is the strong possibility that, under certain circumstances, group membership can exert a protective influence on the individual. Holmes (14) in his studies on tuberculosis in Seattle, for example, has shown that the disease occurs most frequently in "marginal" people, that is, in those individuals deprived of